Don't you have to update batch norm statistics?

[Vladimir]

Hi,

thank you for the implementation!
I can't find update stat operation
from the documentation
update_ops = tf.get_collection(tf.GraphKeys.UPDATE_OPS) with tf.control_dependencies(update_ops): train_op = optimizer.minimize(loss)

https://www.tensorflow.org/api_docs/python/tf/contrib/layers/batch_norm

[Illarion]

Hi,

I've provide additional argument updates_collections=None in call to the batch norm(link) to force the updates in place as per another approach from docs.